The custom dimensions tool is a feature in Lightspeed Analytics that helps you define your own dimensions and naming conventions into data that Lightspeed is sending to Analytics. 

You can learn more about the language of custom dimensions in the Looker syntax reference guide.

There are some key differences between how custom dimensions and table calculations work:

1. Table calculations require their relative data to be on your table. Custom dimensions can read information that is not on your data table. Let's say I want to gather my top-level categories into cohorts of multiple top-level categories. I could use the same formula to create the cohorts, either in a table calculation, or a custom dimension. The visible results are the same:

   

   But if I remove the original top-level category from the table, what happens?

   

   My table calculation won't work anymore, but my custom dimension does.
2. We can pivot and filter on custom dimensions. Table Calculations can only be defined one column at a time. Our Custom Dimensions can pivot so that the data follows the rules that we set automatically...

   

   Also, we can use our custom dimensions as a filter:

   

3. Custom dimension data can exceed the displayed row limits and continue to sort properly. When we have more data than the available rows can display, we will see the yellow Row limit reached bar:

   

   Now if we try to sort using table calculations we get an error:

   

   ...but if sorting by custom dimensions, it will continue to work as any other dimension.

   

4. Custom dimensions can't look at measures, table calculations, or position on the table. The formula builder only sees dimension-specific data. If we look for sales totals in the Edit Custom Dimension tool it won't know where to look:

   

So, some functions may still require you to use table calculations. But we'll find some neat ways to use custom dimensions to make more precise reports.

Check out the video on Introduction to Custom Fields.